What if the next generation of AI products came from studying what's annoying about using AI today?
Artificial intelligence can write, analyze, create images, answer questions and help people build things faster than ever.
That gap between what AI understands you want to accomplish and what you still have to do manually to accomplish it is what we call:
AI WORKFLOW FRICTION.
And we're proposing a company dedicated to eliminating it.
AI Workflow Friction emerged while CiteShare was being rebuilt with AI.
The process became an experiment without us initially realizing it.
CiteShare and AI would develop an idea together. AI could understand what needed to happen next. But moving that understanding into websites and other applications frequently required repetitive human actions.
Those interruptions started producing a question:
If the AI already knows what I'm trying to do, why am I still doing this part manually?
Instead of treating those moments merely as annoyances, AWF proposes treating them as product research.
Every unnecessary click, search, copy, paste, upload and repetitive action could be evidence of a product waiting to be built.
THE FIRST FOUR FRICTIONS
SCREENBRIDGE — “I just took it. Why do I have to find it?”
During AI-assisted work, screenshots can become a language between the human and the AI. A user captures what is happening on another screen so the AI can see the problem and provide the next instruction.
But the user may then have to leave the workflow, find the newly created screenshot among similarly named files, upload it and return to the conversation.
ScreenBridge asks a simpler question:
Why shouldn't the screenshot you just took already be waiting for your AI?
FORMFILL — “You already wrote it. Why am I copying it?”
AI can design an entire contact form, questionnaire or survey—including questions, field types, descriptions and dropdown choices.
Yet the human may still have to copy those instructions into another application one field at a time.
FormFill explores turning AI's understanding directly into structured forms.
AI → FORM
Not:
AI → HUMAN COPY → SWITCH APPLICATION → PASTE → REPEAT
AUTOVOICE — “I'm working. Just tell me what to do next.”
Sometimes the user's eyes and hands need to remain inside the application they're working in.
If the AI has already produced its answer, why should the user repeatedly stop working to request that the response be read aloud?
AutoVoice explores an AI workflow in which responses can automatically be spoken as they arrive—allowing the user to continue working while the AI talks through the next step.
FILEBRIDGE — “You just made it. Where did it go?”
This friction was discovered while creating the very image used to introduce AWF.
AI generated the graphic.
Then the human spent roughly ten minutes trying to find the graphic on his computer so he could put it on the website.
That is AI Workflow Friction.
FileBridge imagines an intelligent recent-files layer organized around human context and recency, rather than cryptic filenames and timestamps.
Instead of searching for:
Screenshot 2026-09-10 at 6.49.27 PM.png
the user should be able to find:
AWF HOMEPAGE IMAGE — created a few minutes ago
or:
LATEST SQUARESPACE SCREENSHOT
The computer remembers the filename.
The human remembers what they were doing.
AWF should bridge the difference.
THE AWF METHOD
AWF doesn't have to guess what AI users need.
It can watch for friction while people accomplish real work:
USE AI → BUILD SOMETHING → HIT FRICTION → DOCUMENT IT → DESIGN A FIX → TEST IT → PRODUCTIZE IT
The frustration becomes evidence.
The workaround becomes research.
And the solution may become a product.
CITESHARE BECOMES THE FIRST LABORATORY
CiteShare is particularly suited to testing this idea because Business Pitches are developed through continuous interaction among ideas, evidence, people and technology.
As CiteShare is built with AI, its own development process can expose problems in human–AI collaboration.
Those problems can become AWF observations.
Those observations can become proposed solutions.
And those solutions can become new CiteShare Business Pitches.
CiteShare builds.
Friction appears.
AWF observes.
A solution emerges.
CiteShare reports the new Business Pitch.
The cycle begins again.
That is Storytelling Serendipity™ applied not only to entrepreneurship—but to the process of working with artificial intelligence itself.
THE QUESTION
AI companies are spending enormous resources making artificial intelligence more capable.
AWF asks a complementary question:
What if we also made using that intelligence less cumbersome?
The smartest AI in the world still loses some of its value every time a human has to perform an unnecessary step between intention and execution.
AWF wants to find those steps.
And eliminate them.
WHAT CAN YOU ADD TO THE PITCH?
Do you regularly work with AI?
What repetitive thing are you still doing that you believe your AI should already be able to do for you?
Tell us about the friction.
Challenge one of these solutions.
Suggest a better one.
Or help us build it.